Article: Construction of multimodal public opinion knowledge graph for regional economic fluctuations Journal: International Journal of Information and Communication Technology (IJICT) 2026 Vol.27 No.43 pp.1 - 28 Abstract: Faced with the lag challenge of traditional macroeconomic indicators in monitoring regional economic fluctuations, this study proposes a multi-modal public opinion knowledge graph construction method that integrates agenda setting and signalling theory. Through cross-modal coding, spatio-temporal fusion and graph structure learning, the massive heterogeneous graph-text public opinion data is transformed into a structured dynamic knowledge graph. Experiments based on the constructed Yangtze River Delta Regional Economy - Multi-modal Public Opinion dataset show that method achieves 78.3% accuracy in predicting the direction of regional quarterly gross domestic product growth rate, which is significantly better than a variety of frontier baseline models (p < 0.01), and performs well in the task of graph link prediction. This study constructs a theory-computational bridge connecting micro social perception and macro economy, and provides an interpretable new paradigm for intelligent research and judgement of regional economy. Inderscience Publishers - linking academia, business and industry through research

Title: Construction of multimodal public opinion knowledge graph for regional economic fluctuations

Authors: Min Qiu

Addresses: School of Economics and Management, Hebi Institute of Engineering and Technology, Henan Polytechnic University, Hebi, 458000, China

Abstract: Faced with the lag challenge of traditional macroeconomic indicators in monitoring regional economic fluctuations, this study proposes a multi-modal public opinion knowledge graph construction method that integrates agenda setting and signalling theory. Through cross-modal coding, spatio-temporal fusion and graph structure learning, the massive heterogeneous graph-text public opinion data is transformed into a structured dynamic knowledge graph. Experiments based on the constructed Yangtze River Delta Regional Economy - Multi-modal Public Opinion dataset show that method achieves 78.3% accuracy in predicting the direction of regional quarterly gross domestic product growth rate, which is significantly better than a variety of frontier baseline models (p < 0.01), and performs well in the task of graph link prediction. This study constructs a theory-computational bridge connecting micro social perception and macro economy, and provides an interpretable new paradigm for intelligent research and judgement of regional economy.

Keywords: regional economic fluctuations; multi-modal public opinion; knowledge graph; signal theory; prediction accuracy.

DOI: 10.1504/IJICT.2026.153384

International Journal of Information and Communication Technology, 2026 Vol.27 No.43, pp.1 - 28

Received: 03 Jan 2026
Accepted: 02 Feb 2026
Published online: 06 May 2026 *

Title: Construction of multimodal public opinion knowledge graph for regional economic fluctuations

Keep up-to-date